This research investigates how national digital strategies affect corporate ESG performance. Utilizing data from Chinese A-share listed companies spanning 2011-2023, we empirically examine the impact of the National Big Data Comprehensive Pilot Zones policy on corporate ESG and its underlying mechanisms using a multi-period difference-in-differences (DID) approach. The findings reveal that: (1) This policy significantly enhances firms' overall ESG performance; (2) The policy effect emerges after a 2-3 year lag, aligning with the "technology absorption → organizational adaptation → governance optimization" transmission path; (3) The policy impact exhibits heterogeneity: responses are more pronounced among firms in the eastern region and non-state-owned enterprises (non-SOEs), while the effect is insignificant for firms in the central and western regions and state-owned enterprises (SOEs). This indicates that decision-making differences stemming from regional resource endowments and corporate ownership structure are key influencing factors. This study provides evidence for understanding the micro-mechanisms through which digital policies drive corporate sustainable development, offering policy implications for optimizing pilot zone construction, promoting coordinated regional development, and precisely guiding enterprises to enhance ESG practices.

The present study explores the forecasting performance of two distinct methods: Historical Simulation (HS) and Monte Carlo (MC). The aforementioned approaches find application in the estimation of VaR of the CBOE Volatility Index (VIX), a benchmark of paramount importance in the assessment of market risk. As financial institutions increasingly rely on VaR models to quantify volatility risk, the choice between computationally efficient but potentially oversimplified HS approaches and MC methods, though more sophisticated, is a key operational decision. This study employs a rolling-window framework with 10-year calibration periods to analyse a three-decade period of VIX data (1990-2023). This methodology is utilised in order to draw comparisons between standard HS, crisis-adjusted HS, and MC simulation incorporating Ornstein-Uhlenbeck processes. The findings reveal that the MC approach attained a statistically significant 12.7% reduction (p <0.01) in 95% VaR forecast errors when compared against HS during normal volatility periods (VIX <25). Furthermore, the MC approach exhibited superior crisis performance, with breach rates deviating 8.2% from theoretical expectations, in contrast to the HS approach, which deviated 31.4%. However, it is important to note that this was achieved at a substantial computational cost of 117 times the processing time (9.4 seconds vs. 0.08 seconds per estimation). The findings of the study provide a decision framework grounded in empirical evidence. It is asserted that the implementation of weighted HS is to be recommended for scenarios involving high-frequency monitoring, and that MC is to be employed for stress testing scenarios. The robustness of the decision framework has been demonstrated to be reliable on multiple occasions, as evidenced by its application during significant market events. These include the 2008 financial crisis and the 2020 pandemic volatility spike. The present text provides practitioners with guidance for the implementation of volatility risk management systems, which has been empirically validated.